Journal of Philosophical Research
Volume 35, 2010
Extremity of Vice and the Character of Evil
Peter Brian Barry
Saginaw Valley State University
ABSTRACT: It is plausible that being an evil person is a matter of having a particularly morally depraved character. I argue
that suffering from extreme moral vices—and not consistently
lacking moral vices, for example—suffices for being evil.
Alternatively, I defend an extremity account concerning evil
personhood against consistency accounts of evil personhood.
After clarifying what it is for vices to be extreme, I note that the
extremity thesis I defend allows that a person could suffer from
both extremely vicious character traits while possessing some
modest virtue as well. By contrast, consistency theses rule out
this possibility by definition. This result does not suggest that extremity accounts are flawed, however, since, as I argue, the thesis
that evil people must lack moral virtue altogether effectively
defines evil people out of existence and prematurely privileges
skepticism about evil personhood. Ultimately, I contend that an
extremity account is most consistent with common intuitions
about putative evil persons as well as plausible assumptions
about aretaic evaluations of character quite generally.
the main fallacy which prevents people from recognizing potential
Hitlers before they have shown their true faces . . . lies in the belief that
a thoroughly destructive and evil man must be a devil—and look his
part; that he must be devoid of any positive quality; that he must bear
the sign of Cain. . . . There is hardly a man who is utterly devoid of any
kindness, of any good intentions. . . . Hence, as long as one believes
that the evil man wears horns, one will not discover an evil man.
—Erich Fromm (1973, 432)
A
I. INTRODUCTION
ny plausible account of evil personhood must take seriously the above observation that many evildoers, even the perpetrators of great atrocities, are in some
26
Peter Brian Barry
respects rather similar to morally decent folk (Morton 2004).1 In what follows, I
articulate and defend an account of evil personhood that is consistent with Fromm’s
observation. In particular, I argue for an “extremity account” of evil personhood
and against “consistency accounts” of evil personhood.2
II. A MODEST PROPOSAL AND TWO THESES ABOUT EVIL
Since Hannah Arendt (1963) introduced the locution, it is common enough to take
the “banality of evil” for granted.3 It is certainly a mistake to suppose that banality
is necessary for being evil: Idi Amin and Charles Manson are hardly boring or ordinary, but they are certainly not exonerated from the charge of being evil on those
grounds. More plausibly, Arendt’s ruminations on the banality of evil are part of
an attempt to reconcile the assumption that Eichmann is both evil and disturbingly
similar to many of we morally decent folk.
Arendt’s discussion of Eichmann is not uncontroversial, and her assumption that Eichmann is evil can be challenged. One problem with supposing that
Eichmann is evil is that while he is surely a morally bad person, he surely could
have been much, much worse; the animus Eichmann lacks, Hitler has in spades.
But if Hitler is worse than Eichmann, it is because he is a morally worse sort of
person than Eichmann—that is, because Hitler has a qualitatively different (and
qualitatively worse) sort of character. A number of philosophers have suggested
that describing someone as “evil” is qualitatively different from describing them
as “bad” or even “very, very bad” (de Wijze 2002; Garcia 2002; Garrard 2002;
Haybron 2002b; Steiner 2002). And a number of philosophers agree that being
evil amounts to having a certain kind of character (Garrard 1998; Haybron 2002b;
McGinn 1997; Perrett 2002). This reasoning is captured in the following thesis I
dub ‘the modest proposal’:
(MP)The evil person is the morally worst sort of person; for all persons, if x is
a morally worse sort of person than y, then y is not evil, and y is evil only
if x is not a morally worse sort of person than y.
(MP) does not imply that there could be only one evil person since it asserts that
a person is not evil unless he or she is the morally worst sort of person. The point
is simply that a necessary condition for being evil is that one could not morally
worse than one is.4 Thus, those who suppose—with Arendt—that Eichmann is
evil must also suppose that Eichmann is at least as bad a sort of person as Hitler
if Hitler is evil.
Suppose, plausibly enough, that ‘evil’ is a superlative such that it is rightly
used to describe only the morally worst sort of person. It would seem that only
the cruelest and meanest and most hateful of us could be evil, since someone who
was less so could be a worse sort of person, and hence is not evil by (MP). Such
reflection supports the extremity thesis:
(ET)A person is evil in virtue of possessing extremely morally vicious states
of character.
Extremity of Vice and the Character of Evil
27
Similarly, if evil persons are the morally worst sort of person, then they surely
must lack morally redeeming qualities. After all, someone possessed of morally
redeeming qualities has a morally worse counterpart who lacks those qualities.
Such reflection supports the consistency thesis:
(CT)A person is evil in virtue of consistently lacking morally virtuous states
of character.
Ultimately, I argue that (ET) is correct and that (CT) is not—that is, that extreme
viciousness suffices for being evil and the consistent absence of moral virtue is
neither necessary nor sufficient for being evil.5
(CT) and (ET) both may appear obviously false. Some putative evil persons seem
to possess morally redeeming features. Arendt herself acknowledges that Eichmann
was not completely without compassion and kindness, noting that he took steps
to help some Jewish relatives to emigrate and that he deeply and publicly regretted shaming “one of his favorite Jews” (Arendt 1963, 46–47). But if evil persons
can possess such virtue, (CT) is false. Arendt also notes that Eichmann fell wildly
short of the expectations of the Nuremburg Court, arguing that despite the efforts
of the prosecution, it was clear he was no “monster” (Arendt 1963, 54). Indeed,
Eichmann’s banality surely rested partly in his lack of animus, precisely the sort
of thing that might be expected of an extremely vicious person. But if evil persons
can lack such vice, (ET) is false.
Much depends on what alleged counter-example to (CT) and (ET) is considered;
if there is good reason to doubt that Eichmann is evil, then he is no counter-example.
Further, reflection on other putative evil persons supports both (CT) and (ET). Consider a particularly extraordinary example of a sociopath: Robert Alton Harris. One
of Harris’s companions on death row suggests that Harris “doesn’t care about life,
he doesn’t care about others, he doesn’t care about himself” (Watson 1993, 130).
Harris’s sister, Barbara, claimed that “every grain of sweetness, pity, and goodness
in him [was] destroyed” (Watson 1993, 134). If these descriptions of Harris can be
taken at face value—if Harris entirely lacked care for himself and others, if every
grain of goodness in him was destroyed—then he would consistently lack virtue as
(CT) suggests. Harris’s lack of care and concern also suggests that he is extremely
callous and cruel and thus suffers from extreme vices as (ET) suggests.
Any account of evil personhood that implies (CT) is a consistency account and
any account of evil personhood that implies (ET) is an extremity account. Both
accounts merit further consideration.
III. CONSISTENCY ACCOUNTS OF EVIL PERSONHOOD
Consistency accounts allow that evil persons may have significant non-moral virtues: just as a cat-burglar might be clever and a philanderer might be conscientious,
an evil person might be resolved or charming or witty (Haybron 2002a). (CT) is
nicely captured in Dan Haybron’s suggestion that evil persons “lack any significant
moral virtues, having no ‘good side’” as it were (Haybron 2002a, 63). Haybron
28
Peter Brian Barry
sometimes characterizes evil persons as those persons lacking morally appropriate
motivation. Using an apt phrase, Haybron claims that being appropriately motivated
by the good is to be “aligned with the good” and that evil persons are “unaligned”
with the good and therefore consistently inappropriately motivated by the good
(Haybron 2002a, 71). It is fairly widely agreed among virtue ethicists that the
virtues are multi-track dispositions, that virtues dispose their agents to perform
certain actions in certain circumstances for certain reasons with certain feelings,
and so forth (Hursthouse 1999). So understood, someone who consistently has no
desire to act rightly when the opportunity arises consistently lacks virtue, since she
consistently lacks dispositions partly constitutive of virtue.
Eve Garrard is also committed to (CT). Garrard suggests that evil persons
suffer from “a profound cognitive defect” (Garrard 2002, 331). In particular, evil
persons suffer from “a total failure to see that certain considerations are [moral]
reasons at all” such that an evil agent “can’t even see that there is a reason of the
most important kind against his action” (Garrard 2002, 330–331). Further, the
pain and suffering of an evil person’s victim “make[s] no claim on him at all that
he can hear” suggesting an “inability to hear the victims’ screams as significant”
(Garrard 1998, 53–54). Quite literally, Garrard’s evil person is unable to perceive
moral reasons.
Garrard’s characterization of evil persons is problematic, however. Garrard’s
claim that evil persons cannot perceive moral reasons suggests that evil people fail
to be receptive to moral reasons. A failure of receptivity involves failed connections
between moral reasons and an agent’s belief about those reasons, as opposed to a
failure of reactivity that involves a failure to translate moral beliefs into morally
appropriate action (Fischer and Ravizza 1998). If, on Garrard’s view, evil people
are not even receptive to moral reasons, then Garrard’s view yields the implausible
result that knowingly engaging in wrongdoing exculpates one from being evil; after
all, for an agent to know that her act is wrong she must, roughly, be aware that
there is some moral reason for her not to act as such. But then she is receptive to
moral reasons and not evil on Garrard’s view.6
Fortunately, a more charitable interpretation of Garrard is available. Garrard
also claims that an evil person is “completely insensitive” to the reason-giving
force of another’s pain (Garrard 2002, 330). One way to be insensitive to another’s
suffering is to be oblivious to it, but an insensitive person might instead be aware
of another’s suffering and disregard its moral significance. That latter sort of insensitivity suggests a failure of reactivity—that is, a failure to appropriately translate
moral beliefs into action. And if virtues are understood as multi-track dispositions,
a consistent failure to be reactive to moral reasons implies a consistent failure to
act rightly for the right reasons, a disposition partly constitutive of virtue. So, on
the most plausible interpretation of her view, Garrard’s evil person consistently
lacks virtue and she is committed to (CT).
Haybron suggests that (CT) can be further refined insofar as he distinguishes
weak and strong varieties of consistency accounts: weak consistency accounts allow
Extremity of Vice and the Character of Evil
29
that an evil person might have some “morally insignificant” virtue—that is, virtue
that does not make a difference to the overall moral quality of her character—while
strong consistency accounts preclude even morally insignificant virtue (Haybron
2002a, 70). So, two different versions of (CT), one weak and one strong, should
be distinguished:
(WCT)A person is evil in virtue of consistently lacking morally significant
virtue.
(SCT)A person is evil in virtue of consistently lacking moral virtue entirely.
(WCT) allows that an evil person may be a bit kind or generous, but not so kind
or generous that the overall moral quality of his character improves. Suppose that
on some few occasions, Hitler exhibited some slight kindness in his relationship
with Eva Braun, acting kindly and for the right reasons with the right feelings, and
so forth, suggesting he possessed some limited virtue. (WCT) does not rule out
the possibility that this Hitler is still evil, perhaps because his slight virtue makes
no difference to the overall moral quality of his character, though he is clearly not
evil on (SCT).
Interestingly, Haybron himself rejects (WCT) noting that mere ne’er-dowells—persons who genuinely care about the good of others but who are never
adequately motivated to act accordingly—lack significant virtue but are wrongly
regarded as evil (Haybron 2002a). And that seems right, suggesting that (SCT) is
the most plausible thesis available to proponents of consistency accounts. Thus,
in what follows, I suppose that (SCT) is the definitive consistency thesis and that
consistency accounts fail or succeed on its merits.
IV. EXTREMITY ACCOUNTS OF EVIL PERSONHOOD
Recall (ET): a person is evil in virtue of suffering from extremely vicious states of
character. No philosopher that I am aware of has explicitly defended (ET); hence,
it is somewhat unclear just what extreme viciousness amounts to. Fortunately,
proponents of (ET) can take advantage of recent work on the virtues. Thomas
Hurka suggests that there are “degrees” of virtue and vice and that their degree is
a function of both their intensity and the value of their object (Hurka 2001, 58).
Proponents of (ET) can thus plausibly explicate the extreme degree of some vice
either in terms of its intensity or in terms of the value of its object.
First, consider the suggestion that vices are extreme if they are sufficiently
“intense”: the more intense, the more extreme. The intensity of an attitude is most
plausibly understood in terms of its capacity to motivate its agent (Hurka 2001).
Other things being equal, the more a character trait disposes its agent to perform
some requisite action, the more intense that character trait. Extreme vices would
thus be those vices that greatly dispose their agent to perform some requisite
wrong action. It is this first sense of ‘extreme’ that Haybron seems to have in
mind when he claims that an evil person who is cruel will tend to have “cruel
propensities to the highest degree” since cruel propensities will most obviously
30
Peter Brian Barry
manifest themselves in cruel actions (Haybron 2002a, 63). So, other things being
equal, if Rosencrantz greatly and regularly wants to inflict suffering on Hamlet
while Guildenstern weakly wants to do so, then Rosencrantz’s cruelty is more
extreme in this first sense.
Following Hurka, however, vices can be extreme if their object is sufficiently
disvaluable: the more morally disvaluable the object, the more extreme the vice.
Other things being equal, if Rosencrantz is unjust while Guildenstern is merely
apathetic, Rosencrantz suffers from a more extreme vice in this second sense
since his vice has a morally worse object. The worst moral vices—those that are
extremely bad—would be those that have the most morally disvaluable objects.7
Vices are extreme in this second sense insofar as they are correlated with especially
disvaluable objects. The objects of both malevolence and rudeness are both morally
disvaluable: the object of malevolence is some state of affairs in which undeserving persons are harmed and the object of rudeness is some state of affairs in which
persons are offended. But surely the object of malevolence is of great moral disvalue
and greater moral disvalue than the object of rudeness. Thus, malevolence, but not
rudeness, is an extreme sort of vice.
It is of course possible that vices might be extreme in one sense but not the
other; someone might be malicious but never motivated to act accordingly and
someone might be merely rude though greatly motivated to act accordingly. Probably, a person who suffers from vices that are extreme in only one of the above
two senses is not evil. (ET) is most plausible when the vices of the evil person are
extreme in both senses—that is, when she suffers from the morally worst sort of
vices to significant degrees.
A plausible consequence of (ET) should be noted. It would be a mistake to
identify the evil person with someone disposed to act wrongly, since someone
might act wrongfully while feeling shame or regret or remorse for her actions,
emotions that are inconsistent with having an evil character. But if virtues and vices
are understood as multi-track dispositions, evil persons with extreme vices will
not just be disposed to act wrongly but also to act for wrong reasons, to have the
wrong feelings and so forth. Further, it is probably the case that virtues and vices
are grouped together in important ways. I discuss this point further below, but it is
rather plausible to suppose that someone who is unjust will also be miserly, that
someone who is malicious will also be callous, and so forth. Thus, the possession
of some extreme vice will tend to bring with it other moral vices with similar
constitutive dispositions. Insofar as evil persons suffer from extreme vices, their
characters will likely be well-rounded with other morally dubious character traits
as well, which is probably just what we expect of evil persons.
(ET) also helps to make sense of the above suggestion that Eichmann is not
clearly evil because Hitler seems to be a morally worse sort of person than Eichmann. If Hitler is malicious and hateful while Eichmann is “merely” callous and
selfish, then Hitler’s vices are extreme in a way that Eichmann’s were not because
Eichmann’s vices are not among the morally worst. If Hitler has a qualitatively
Extremity of Vice and the Character of Evil
31
worse moral character than Eichmann, then, by (MP), Eichmann is not evil whatever else is wrong with him.
(ET) seems well supported by examining concrete examples of putative evil
persons. Focus again on Robert Alton Harris. Harris, accompanied by his brother
Daniel, killed two sixteen year-old youths after stealing their car, shooting one in
the back and chasing the other down a hill, shooting him four times. While Daniel
reacted with the expected shock and horror, Robert contented himself with eating the slain youths’ hamburgers and apple turnovers, laughing as Daniel became
nauseated. Harris then proposed that he and Daniel pose as police officers and
inform the parents of their victims that their sons had been killed, laughing while
recounting how he “really blew that guy’s brains out” (Watson 1993, 131–133).
Harris’s indifference and callousness was not slight, but grave. Insofar as Harris
seems to have relished hunting his youthful victim, he is not just a bit malevolent
and sadistic, but very much so. Insofar as Harris suffers from some of the morally
worst sort of vices and suffers from them to significant degrees, Harris suffers from
extreme moral vices.
(ET) seems to get other cases right; in particular, while the proponent of (ET)
can agree that paradigmatic examples of evil people—like Hitler and Harris—really
are evil, she can be genuinely ambivalent about difficult and controversial cases.
For example, some accounts of Eichmann suggest a rather different sort of person,
still callous but not to the mythical degree suggested by Arendt (Cesarani 2004).
(ET) allows for ambivalence about whether or not Eichmann really is evil precisely
because it is unclear just how intense his vices are. Since I take it that it is generally a virtue of an analysis that it does not artificially do away with controversial
cases, I suppose that it is a virtue of (ET) that it permits sincere ambivalence with
respect to controversial cases of putatively evil people.
V. TWO CHARACTER STUDIES
While I argue that (ET) gets cases right, Haybron argues that (ET) actually gets
cases wrong and that (CT) fares rather better. Haybron’s arguments depend upon
two character studies, both of which I consider below.
Haybron first considers Tony Soprano, fictional New Jersey Mafioso. Soprano
plausibly suffers from extreme vices as suggested by his morally wrong acts: Soprano cheats on his wife Carmela repeatedly; he goes to the hospital with the intent
of smothering his mother to death; on several occasions, he takes advantage of his
friends in his usury business; he assaults and batters and kills for the sake of la
Cosa Nostra, twice killing individuals he previously claimed were his best friend.
Still, Soprano often enough displays a good side: on a tour of college campuses,
Soprano shows genuine warmth and love for his daughter (and later kills a mob
informant he stumbles upon); he seems to genuinely regret some of his behavior, at
least for a while; he is genuinely repulsed by some actions of others. If Soprano’s
immoral actions imply he has extreme vices, his right actions suggest that he is
not entirely without virtue.
32
Peter Brian Barry
Haybron next considers Chad, protagonist of Neil LaBute’s film In the Company of Men. While Soprano’s actions suggest that he is not without virtue, Chad
does nothing to suggest that he has any virtue whatsoever. Chad and his spineless
co-worker, Howard, charm a naive young deaf woman with the express purpose
of making her think that she has betrayed them both. Chad and Howard intend to
leave her as revenge for the abuse each has suffered from their respective girlfriends.
Chad prematurely informs her of their plan and takes relish in her suffering. Howard
ultimately loses his job to Chad, partly through Chad’s sabotage. In the final scenes
of the movie, it is revealed that Chad and his girlfriend never separated and that he
merely wanted to “really hurt someone.” Along the way, he humiliates and abuses
fellow co-workers and takes every opportunity to engage in whatever misogyny
he can. Chad appears morally corrupt to the bone, utterly lacking a morally good
side, consistently unaligned with the good.
Given Soprano’s extreme vices, he should be evil according to (ET). But that,
Haybron thinks, is the problem, since Soprano is not credibly viewed as evil (Haybron 2002a). Haybron offers two arguments for this conclusion. He claims that,
first, morally decent people are disappointed in Soprano at times; however, evil
persons do not disappoint morally decent people since no one expects evil people
to act any better. Of course, evil persons might disappoint if it was mistakenly
thought that they were good persons—the neighbors of serial killers are often
enough surprised to learn that their neighbor who seemed “so normal” was capable
of murder. Still, upon learning that they are evil, little that they do disappoints.
Second, the members of Soprano’s immediate family are basically good people
whom Soprano fits in with reasonably well; however, evil people do not fit in with
good people (Haybron 2002a).
I suspect that neither of Haybron’s arguments is sound. Focus first on the suggestion that evil people do not fit in well with morally decent people. Whether
evil people do not fit in well with morally decent people surely must be determined empirically; it is difficult to see how it could be shown a priori that they
could not get along. However, it is easy to explain the hope that they could not.
Perhaps getting along well with evil persons comes too close to sanctioning or
sympathizing with them. The insistence that decent people could not get along
well with evil persons might then be an attempt to avoid cognitive dissonance,
to avoid admitting that willing consort with evil persons ever happens. Bur precisely for those reasons, the thesis that decent people do not fit in well with evil
people is suspect.
In any case, it seems rather unlikely that Soprano does fit in well with his family: besides cheating on his wife and threatening to kill his mother, he fights with
his uncle Junior and his sister Janice, kills his cousin Tony and his nephew Christopher, and alienates his children. It is also far from clear that Soprano’s family is
composed of basically good people: Carmela is every bit as ruthless as Soprano
himself and she rarely shows regret for the lavish lifestyle she lives, knowing full
well where the money comes from; no one at the funeral of Soprano’s mother
Extremity of Vice and the Character of Evil
33
has a kind word to say about her; Junior is a narcissist who tries to have Soprano
killed; Christopher is a racist, a murderer and a drug-addict who beats his fiancé.
Perhaps it is not surprising Soprano fits in so well with his family, if he does; we
could hardly imagine a more morally bankrupt lot.
It is also worth wondering whether evil persons never disappoint morally decent
persons: if evil people are not without virtue, they might. Arguably, part of what
makes Soprano such a terrible person is the fact that he could have acted better
than he did, as evidenced by his right actions. Familiarly, most incompatibilists
about moral responsibility insist that a person is an apt candidate for reactive attitudes like blame and resentment only if he could have done otherwise. If calling
someone “evil” is a means for exercising the reactive attitudes then evil persons
would have to be capable of doing otherwise and acting better: Soprano’s modest
virtue does not disqualify him from being evil, but would rather be essential to it.
To be sure, my case against (SCT) does not rest on incompatibilist intuitions, but
those intuitions are not clearly implausible.
It is also far from clear that Chad fails to suffer from extreme vices. Chad
suffers from some of the morally worst vices, including malice and malevolence,
and suffers them in spades. Haybron doubts that Chad is extremely vicious
as required by (ET), given that “whether he frequently does serious harm is
questionable” and that Chad “does not represent the kind of danger that a Tony
Soprano represents” (Haybron 2002a, 69). However, (ET) does not imply that an
evil person must frequently cause serious harm or represent a significant danger,
even if it requires that he is disposed to do so. And while it is plausible to suppose that evil persons will tend to perform the morally worst sorts of actions, it
does not follow that they must or even that they will do so particularly often. So
neither Soprano nor Chad is a clear counter-example to (ET) and Chad does not
singularly confirm (SCT).8
VI. UNITY OF THE VIRTUES AND VICES
By definition, (SCT) rules out the possibility that someone could both be evil and
possess some virtue. But since (ET) only demands that evil people suffer from
extreme moral vice, it does not rule out the possibility that someone could be both
evil and somewhat virtuous.9 So, (ET), unlike (SCT), is consistent with Fromm’s
caution against supposing that evil persons must be devoid of any positive quality, utterly lacking in kindness and good intentions. Proponents of consistency
accounts might dismiss Fromm’s warning, but only at a cost. (SCT) prematurely
lends credence to evil-skepticism—that is, the thesis that either no one is actually
evil or that very, very few actual people are (Russell 2006). Admittedly, it would
beg the question against evil-skepticism to suppose that many actual evil persons
exist, but since very few actual persons lack moral virtue entirely (SCT) all but
reduces evil personhood to a mere conceptual possibility. Even Haybron supposes
that there is something dubious about effectively defining evil persons out of existence (Haybron 2002b).
34
Peter Brian Barry
However, on a familiar line of reasoning that goes back to Plato and Aristotle,
moral virtue and vice are mutually incompatible so that this alleged advantage of
(ET) is incoherent. Worse, the fact that it implies something incoherent is a reason
to dismiss (ET) altogether.10 So, some reflection on the relationship of the virtues
and vices is called for if (ET) is not to be dismissed as implausible.
I propose to consider the assumption that moral virtue and vice are incompatible.
On this view of things, a person cannot be courageous unless she is also, inter alia,
prudent and generous and temperate; generally, possession of any one virtue implies
possession of them all. Call this the “unity of the virtues” thesis or (UV):
(UV)A person can possess some moral virtue just in case she possesses every
other moral virtue as well; failure to possess any moral virtue implies a
failure to possess every other.
If (UV) is correct, then no one is just but impatient, honest but unkind, and so
forth. (UV) still finds adherents (Annas 1993; McDowell 1998) though it is often
enough rejected (Flanagan 1991; Foot 1983; Hurka 2001; Williams 1985). While it
is beyond the scope of this paper to consider its merits, note that the proponent of
(SCT) had best not embrace (UV). The model of the virtuous person suggested by
(UV) is neither easily nor commonly realized: most of us suffer from some vices
and, by (UV), most of us would have to lack moral virtue altogether. But then, by
(SCT), most of us are evil and that result is simply implausible.
Still, whatever reasons there are to reject (UV), it is not terribly plausible that
the virtues are entirely disunited—that someone could be genuinely kind but also
unjust, stingy, intemperate, and cruel, for example. An argument that the virtues are
not entirely disunited is available if we take seriously the thought that the virtues
are grounded in multi-track dispositional states: some dispositions are bound to be
constituents of more than one virtue. For example, the disposition to act in ways
that benefit others is tied up with kindness but also with generosity and justice, and
reliably telling the truth to others is tied up with honesty but also with helpfulness
and magnanimity. In broad strokes, some virtues are consistently correlated—and
hence, not entirely disunited—with others because they share constitutive dispositional states. Similar remarks apply to the vices: some dispositions are constituents
of more than one vice and thus some particular vices are consistently correlated
with others.
The reasoning of the above paragraph suggests that some virtues and vices are
incompatible since their constitutive dispositions are incompatible with one another:
if kindness is partly constituted by a disposition to help others while stinginess is
partly constituted by the absence of such a disposition, then kindness and stinginess are incompatible, for example. But not all such dispositions are incompatible;
even if courage and timidity are constituted by clearly incompatible dispositions,
courage and patience are not. But then it should be possible for the same person to
suffer from both extreme vice and modest virtue—say, extreme timidity and modest
patience—so long as those they are not constituted by incompatible dispositions.
Extremity of Vice and the Character of Evil
35
Another argument suggests that modest virtue and extreme vice are compatible.
Interestingly, while he defends (UV), Aristotle also seems to claim that a person
might be virtuous in his own affairs without being virtuous—that is, just—in his
affairs with others (Aristotle 1999, 69).11 Virtues and vices might be limited to
particular domains, where a domain is understood as an area of important practical concern that can be psychologically isolated from other practical concerns
(Badhwar 1996). For example, a person might be generous in his dealings with
family and friends but not in his dealings with strangers and thereby possess and
exercise generosity in one domain but not another. Now, the dispositions that are
constitutive of any particular state of character may be more or less fine-grained:
a person may count as being generous partly in virtue of being disposed to help
others or in virtue of being disposed to help others in great need; a patient person
may be disposed to tolerate others generally or children in particular. So long as
various domains can be compartmentalized from one another—that is, so long as
it is psychologically possible to care especially about one area of concern but not
another—it may well be possible to possess enough of the dispositions constitutive of some virtue that one could rightly be said to be generous or kind or patient
within a domain without being generous or kind or patient tout court.
What, then, to say of someone who exercises kindness towards his fellow
citizens but appears strongly disposed to cruelty towards foreigners? In everyday
moral parlance, we might say that he is both generous and cruel, shorthand for
the more complicated (and coherent) assertion that he is both generous within a
domain and cruel within a domain. It is strained to say that he is not kind at all,
given his dispositions to act and feel rightly towards his fellow countrymen. But
if his xenophobia leads him to be exceptionally cruel, it is surely reasonable to
suppose that his kindness is modest, at best. But then he plausibly possesses both
extreme vice and modest moral virtue.
If the arguments of this section are correct, there is no implausibility in claiming
that someone could suffer from extreme vice and modest virtue. But (ET) is more
than simply plausible; there are good reasons to suppose it is true.
VII. FOUR ARGUMENTS FOR (ET)
At least four arguments imply that (ET) is correct, that (SCT) is false, or both. I
consider them individually.
A. The No-Motive Argument
It is rather plausible to suppose that evil persons are disposed to act wrongly; any
plausible account of evil personhood will clarify which psychological properties
and states are constitutive of evil character in virtue of which evildoing is no accident (Card 2002). But then any account of evil personhood that fails to account
for the fact that evil persons are so motivated is incomplete at best and implausible
at worst.
36
Peter Brian Barry
If (ET) is correct, then it is no surprise that evil persons are motivated to act
wrongly: being evil is a function of suffering from moral vices to rather significant
degrees according to (ET), and being disposed to act wrongly is partly constitutive
of having those vices. But if (SCT) is correct, evil people may well have no motive
to act wrongly. On Haybron’s account, evil persons might be unaligned with the
good but not “aligned with the bad”—that is, they need not possess any motivation
to cause harm or wreak havoc or otherwise act wrongly. Similarly, Garrard’s evil
person might be completely insensitive to moral reasons for action, but not “sensitive” in the least to immoral reasons for action—that is, they need not be receptive
or reactive in the least to reasons to act wrongly. Generally, if (SCT) is true, being disposed to act wrongly is extrinsic to evil personhood—at best a contingent
matter—such that it is an accident that evil people engage in wrongdoing.
In fairness, Haybron is aware of this objection. Against Garrard, he objects that
her account “seems not to allow for more virulent strains of wickedness; those for
whom the undeserved suffering of others is not silenced, but is rather a source of
attraction” (Haybron 2002b, 281–282). Indeed, Haybron allows that “Evil persons
could . . . be actively opposed to the good” (Haybron 2002a, 71). Obviously, such
amendments go beyond anything like (SCT). But such amendments push the consistency theorist towards (ET). For it is unlikely that evil persons are only slightly
motivated to perform actions that are somewhat wrong; more likely, they will be
significantly disposed to perform morally wrong actions and particularly horrible
ones at that. But a stronger desire to perform terribly morally wrong actions at
least suggests an agent who suffers from the morally worst vices to significant
degrees—that is, that she suffers from extreme vice.
B. The Difference Argument
It is not only evil persons who suffer from vices; many merely morally bad people—
people who are not quite evil, however bad they are—suffer from them, and so do
many of we morally decent folk. Something distinguishes evil persons from merely
morally bad people, since merely morally bad people could be worse and, therefore,
are not evil by (MP). A plausible account of evil personhood explains why and
how the character of the evil person is qualitatively worse than the character of the
morally bad person, for example.
(ET) can account for the difference. Evil people are not merely vicious according to (ET): they suffer from the morally worst sort of vices, vices like maliciousness and cruelty. They may also be slothful or rude but those are not the vices that
justify calling them “evil.” Of course, morally bad people might also be malicious
or cruel, but probably not to the degree that evil persons are; evil persons are not
just somewhat malicious or cruel but very much so according to (ET). By contrast,
morally bad persons and morally decent persons may suffer from vices that are
extreme in one of the two relevant senses but not both. The fact that (ET) allows
for a principled way to distinguish between morally different sorts of persons is a
reason to accept it.
Extremity of Vice and the Character of Evil
37
C. The No-Change Argument
(SCT) implies that if a person has any moral virtue then she is not evil. Thus, according to (SCT), if a formerly evil person comes to acquire some slight moral virtue,
then she is no longer evil whatever else is wrong with her. So, (SCT) implies that
a slight change in an evil person’s aretaic properties yields a significant change in
the overall moral assessment of her character.
However, it is not generally true that slight changes in a person’s aretaic properties yield significant changes in her overall character. If a moral saint—more on
them in a moment—becomes just a bit more caring or selfless, her overall moral
character might not improve if only because she is already such a morally worthy
person. Similarly, if she acquires just a bit of impatience or rudeness, her overall
moral character still might not change if only because she is already so morally
good: acquiring just a slight bit of a comparatively trivial vice does not “poison
the well,” as it were.
For similar reasons the overall moral assessment of an evil person need not
change simply because she comes to acquire a bit of moral virtue. Individual readers
will have to consult their intuitions here, but imagine as vicious a person as can be
imagined—as spiteful, hateful, malicious, malevolent, uncaring, unjust, callous,
and cruel a person as can be conceived—with absolutely no morally redeeming
qualities whatsoever; call him V-1. Then imagine that such a person acquires just
a bit of virtue; call him V-2. By (SCT), that person is no longer evil, whatever else
is wrong with him or her. Thus, by (MP), the consistency theorist must say that V-1
is a morally worse sort of person than V-2 since she supposes that the former, but
not the latter, is evil. But not only is it not generally true that a slight change in a
person’s aretaic properties must yield a significant change in the overall moral quality of her character, it is not true in this case. It is absurd that such an exceptionally
vicious person is exculpated from the charge of being evil by becoming just a bit
honest or patient, for example. (SCT) implausibly suggests that a significant change
in a person’s character emerges when there is no such change; if the acquisition of
just a bit of moral vice need not poison the well, acquisition of moral virtue need
not clear the water either.
(ET) fares better, by contrast. Since both V-1 and V-2 suffer from extreme vice,
(ET) implies that they are both evil, whatever slight moral virtue they also have:
their extreme vices are sufficient to poison the well and justify describing them
as evil.
D. The Mirror Argument
I offer this last argument in favor of (ET) more tentatively. (ET) allows us to make
sense of a thesis that is sometimes defended in the literature on evil, namely, that
the character of the evil person mirrors the character of the moral saint—that is,
that the morally worst sort of person and the morally best sort of person mirror one
another. Call this the “mirror thesis” (Barry 2009). No small number of philosophers
38
Peter Brian Barry
endorse something like the mirror thesis: David McNaughton (1988, 135) suggests
that the evil person is “the mirror image of the virtuous person”; Colin McGinn
(1997, 62–63) proposes, roughly, that while morally good persons derive pleasure
from the pleasure of other morally good persons and pain from their pain, evil
persons derive pleasure from the pain of others and pain from their pleasure; Hillel Steiner (2002, 185) claims that the actions of evil persons are the “negative
counterparts” of the actions of moral saints; even Haybron (2002b, 274) suggests
that the moral saint is “the positive counterpart to the evil person,” while Garrard
(1998, 49; 2002, 329) suggests that the evil person occupies a space “at the other
end of the moral spectrum” from the moral saint. The problem is that the mirror
thesis is a metaphorical characterization that needs to be rendered literal. I submit
that proponents of (ET) can offer a plausible literal rendering of the mirror thesis
while proponents of (CT) cannot. So, if there is any plausibility to the mirror thesis,
then the fact that (ET) suggests a plausible literal rendering speaks in its favor.
While I have been considering extremity and consistency theses about evil
persons, there are similarly extremity and consistency theses about moral saints
as well. Initially, consider a consistency thesis concerning moral sainthood that
parallels (CT):
(CTS)A person is a moral saint in virtue of consistently lacking morally vicious states of character.
Haybron appears to endorse something like (CTS) insofar as he claims that the
moral saint “has no significant moral flaws or vices, but is perfectly virtuous or
nearly so” (Haybron 2002b, 274). (CTS) thus resembles (CT) insofar as it implies
that the absence of vice suffices for being a moral saint—that is, the morally best
sort of person.12 Consider also an extremity thesis concerning moral sainthood
that parallels (ET):
(ETS) Moral sainthood consists in the possession of extremely morally virtuous states of character.
(ETS) is most plausible if ‘extreme’ is understood unequivocally such that it has the
same meaning as it does in (ET). Thus, the virtues of moral saints, so understood, are
extreme in the sense of being the morally best sort of virtues and in virtue of being
possessed to significant degrees. (ETS) resembles (ET) insofar as it implies that
the possession of extreme virtue suffices for being a moral saint. Susan Wolf seems
at least sympathetic with (ETS) in her seminal discussion of moral sainthood. For
example, Wolf claims that that a moral saint “will have the standard moral virtues
to a non-standard degree” and to “an extreme degree” (Wolf 1982, 421). But Wolf
also claims that “a necessary condition of moral sainthood would be that one’s life
be dominated by a commitment to improving the welfare of others or of society as
a whole” (Wolf 1982, 419–420). A person possessing such a commitment is, inter
alia, generous and just and these are surely among the morally best sort of virtues.
And since her life is “dominated” by such a commitment, she is not just somewhat
generous and just but very much so.
Extremity of Vice and the Character of Evil
39
Note also that while (CTS) precludes the possibility, (ETS) allows that even
moral saints might suffer from some vices. And that is one problem for (CTS).
For while there might be disagreement about just who counts as the morally best
sort of person, most if not all putative moral saints suffer from not insignificant
moral failings and corresponding vices: Martin Luther King was a philanderer;
Mother Teresa confessed that for some period of time she lost her faith and on
occasion declared suspect motives for caring for the hungry; Oskar Schindler’s
motives may not have been entirely pure; Gandhi was estranged from his family
and arguably engaged in exploitive sexual relations with women to test his ascetic
resolve; Socrates neglected his family; Jimmy Carter, a deeply pious and religious
man, admitted to having lust in his heart, and so forth (Flanagan, 1991; Hamilton
1999). It is likely the case that (CTS) reduces the moral saint to a mere conceptual
possibility, just as (SCT) all but defines evil persons out of existence.
There are other objections to (CTS)—objections similar to those against (SCT)—
that suggest that it should be rejected: (CTS) surprisingly allows that moral saints
lack virtue altogether since the consistent absence of vice suffices for moral sainthood; it does not suppose that moral saints are the least bit motivated to act rightly
so that being disposed to help others or pursue justice is an extrinsic and contingent
matter for moral sainthood; it implausibly implies that a perfectly virtuous person
is no longer a moral saint if she becomes just a bit impatient or timid. As such,
(CTS) needs to be rejected as an account of moral sainthood.
By contrast, (ETS) fares rather better. Putative moral saints surely possesses
and exhibit extreme virtue: King was not just somewhat courageous but extremely
so; Mother Theresa was not just somewhat compassionate but extremely so, and
so forth. Even Haybron allows that it is the “profound compassion” of the moral
saint that impresses us (Haybron 2002b, 274). This observation also suggests an
argument in favor of (ETS), one that resembles an argument for (ET). Moral saints
should be distinguished from merely morally decent persons. Morally decent
people certainly possess some virtue and they may even possesses some of the
morally best virtues, but while they may be somewhat compassionate and loyal
and courageous and so forth they are surely not virtuous to the degree that moral
saints are. So, (ETS) suggests a principled qualitative difference between decent
people and moral saints: the virtues of moral saints are extreme in a way that those
of merely decent people are not. Further, (ETS) suggests a reason to suppose that
moral saints are rightly regarded as the morally best sort of person even if they suffer from comparatively slight moral vice: since the virtues of moral saints include
the morally best virtues possessed to significant degrees, they dwarf and outweigh
their comparatively slight moral vices rendering them morally insignificant to the
overall evaluation of their moral characters.
Now, if, as I have argued, (ET) is a plausible account of evil personhood and
(ETS) is a plausible account of moral sainthood, then there is a clear sense in which
evil persons and moral saints mirror one another: the character of the evil person
is marked by extremely vicious states of character while the character of the moral
40
Peter Brian Barry
saint is marked by extremely virtuous states of character. So, (ET) supports a literal
reading of the mirror thesis. No similar literal reading is available to proponents of
(CT). So, if the mirror thesis seems at all plausible, then the fact that (ET) allows
us to make literal sense of it while (CT) does not is at least some reason to embrace
an extremity account of evil personhood.
VIII. CONCLUSION
It is perhaps no surprise to learn that evil persons suffer from vices, but if my arguments above are sound, then it is extreme viciousness that suffices for being evil.
Whether any actual people are evil remains an open question. For all that has been
said, it may be that no one actually possess a sufficiently vicious character that he
or she deserves to be described as evil and whether anyone does possess such a
depraved character is a matter to be determined empirically. Still, (ET) represents
an important start in enquiry into evil personhood and it suggests that Fromm was
right: if being evil does not require curved horns and cloven hooves, neither does
it preclude some kindness and good intentions.
ENDNOTES
1.A number of social psychologists are at pains to make this point. See Staub 1989; Waller,
2002; Zimbardo, 2007.
2. Dan Haybron introduces the terms “extremity account” and “consistency account” in
Haybron 2002a, 63.
3. Stephen Whitfield boldly declares “Arendt had coined a phrase that has become an
inescapable part of the language of modern intellectual life” in Whitfield 1981, 471.
4. My suggestion that the evil person just is the morally worst sort of person is at least
sometimes met with skepticism. One anonymous reviewer fairly captures the typical concern
in the following way: “just as it is always possible to be or do something worse than one
is or has done, it is always possible to be a worse sort of person. . . . I think that you could
always add something to this sort which would make it even worse.” By contrast, I do not
think that it is always possible to be or do something worse than one is or has done. I have the
intuition, and I suspect that many others do too, that our moral evaluations of some actions
and persons simply “bottom out.” For example, it strikes me as plausible that there are some
actions and events that are so harmful that we would not judge them to be still worse if we
learned that they actually caused somewhat more harm; it is not uncommon to talk about
the Holocaust this way. Similarly, it strikes me as plausible to suppose that some people are
so vicious that that we would not judge them to be still worse if we learned that they were
actually somewhat more vicious than we realized; it is not uncommon to talk about Hitler
in this way. I pursue this matter further below.
5. Henceforth, I use ‘virtue’ and ‘vice’ to refer to the moral virtues and vices unless
otherwise noted.
6. Such an agent may still be blameworthy, as one reviewer correctly notes, but that does
not affect my objection here: perhaps genuinely feeling regret or remorse or shame at one’s
Extremity of Vice and the Character of Evil
41
wrongdoing implies that a person is not evil but surely not deliberately and knowingly acting
wrongly in the first place.
7. Hurka suggests, plausibly enough, that taking pleasure in another’s agonizing pain is
worse than taking pleasure in another’s mild pain, presumably because another’s agonizing
pain is a comparatively worse object. See Hurka 2001, 58.
8.To clarify, I do not suppose that I have established that Soprano is evil, only that he
does not clearly fail to qualify as evil. Perhaps Soprano’s vices are sufficiently extreme in
the relevant senses; if so, then he is evil. But perhaps Soprano’s modest moral virtues do
sufficiently mitigate his viciousness; if so, then he is not evil.
9.Thanks to an anonymous reviewer for forcing me to clarify here.
10.As an anonymous reviewer suggested to me.
11.Aristotle 1999, 1129b31–1130a14.
12.A number of medieval philosophers, including Anselm and Augustine, appear to endorse
a consistency account of moral sainthood insofar as they claim that the most saintly and
blessed are those who are non posse peccare: not able to sin. However, it is unclear that
the doctrine of non posse peccare ranges over human beings like us rather than divine or
blessed creatures.
Bibliography
Annas, Julia. 1993. The Morality of Happiness. Oxford: Oxford University Press.
Arendt, Hannah. 1963. Eichmann in Jerusalem: A Report on the Banality of Evil. New
York: The Viking Press.
Aristotle. 1999. The Nicomachean Ethics. Translated, with Introduction, Notes, and Glossary
by Terence Irwin. Indianapolis, IN: Hackett Publishing.
Badhwar, Neera Kapur. 1996. “The Limited Unity of Virtue.” Noûs 30: 306–329.
Barry, Peter Brian. 2009. “Moral Monsters, Moral Saints, and the Mirror Thesis.” American
Philosophical Quarterly 46: 163–176.
Card, Claudia. 2002. The Atrocity Paradigm. Oxford: Oxford University Press.
Cesarani, David. 2004. Becoming Eichmann. United States: DaCapo Press.
de Wijze, Stephen. 2002. “Defining Evil: Insights from the Problem of ‘Dirty Hands.’” The
Monist 85: 210–238.
Fischer, John Martin, and Mark Ravizza. 1998. Responsibility and Control: A Theory of
Moral Responsibility. Oxford: Oxford University Press.
Flanagan, Owen. 1991. Varieties of Moral Personality: Ethics and Psychological Realism.
Cambridge: Harvard University Press.
Foot, Philippa. 1983. “Moral Reason and Moral Dilemma.” Journal of Philosophy 80:
379–398.
Fromm, Erich. 1973. The Anatomy of Human Destructiveness. New York: Henry Holt and
Company.
Garcia, Ernesto. 2002. “A Kantian Theory of Evil.” The Monist 85: 194–209.
Garrard, Eve. 1998. “The Nature of Evil.” Philosophical Explorations 1: 43–60.
42
Peter Brian Barry
———. 2002. “Evil as an Explanatory Concept.” The Monist 85: 320–336.
Hamilton, Christopher. 1999. “The Nature of Evil: A Reply to Garrard.” Philosophical
Explorations 2: 122–138.
Haybron, Daniel. 2002a. “Consistency of Character and the Character of Evil.” In Earth’s
Abominations: Philosophical Studies of Evil, ed. Daniel Haybron New York: Rodopi,
63–78.
———. 2002b. “Moral Monsters and Saints.” The Monist 85: 260–284.
Hurka, Thomas. 2001. Virtue, Value, and Vice. Oxford: Oxford University Press.
Hursthouse, Rosalind. 1999. On Virtue Ethics. Oxford: Oxford University Press.
McDowell, John. 1998. “Virtue and Reason.” In Mind, Value, and Reality, ed. John McDowell.
Cambridge: Harvard University Press, 50–73.
McGinn, Colin. 1997. Evil, Ethics, and Fiction. New York: Oxford University Press.
McNaughton, David. 1988. Moral Vision. Oxford: Blackwell Publishers.
Morton, Adam. 2004. On Evil. New York: Routledge.
Russell, Luke. 2006. “Evil-Revivalism Versus Evil-Skepticism.” The Journal of Value
Inquiry 40: 89–105.
Perrett, Roy. 2002. “Evil and Human Nature.” The Monist 85: 304–319.
Staub, Ervin. 1989. The Roots of Evil. Cambridge: Cambridge University Press.
Steiner, Hillel. 2002. “Calibrating Evil.” The Monist 85: 183–193.
Waller, James. 2002. Becoming Evil: How Ordinary People Commit Genocide and Mass
Killing. Oxford: Oxford University Press.
Watson, Gary. 1993. “Responsibility and the Limits of Evil: Variations on a Strawsonian
Theme.” In Perspectives on Moral Responsibility, ed. John Martin Fischer and Mark
Ravizza. Ithaca: Cornell University Press, 119–148.
Whitfield, Stephen. 1981. “Hannah Arendt and the Banality of Evil.” The History Teacher
14: 469–477.
Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard
University Press.
Wolf, Susan. 1982. “Moral Saints.” The Journal of Philosophy 79: 419–439.
Zimbardo, Philip. 2007. The Lucifer Effect: Understanding How Good People Turn Evil.
New York: Random House Publishing.